#Importing Libraries for ML

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import missingno as msno
from scipy.stats import ttest_ind

from sklearn.svm import SVC
from sklearn.neighbors import KNeighborsClassifier
from sklearn.ensemble import RandomForestClassifier
#from xgboost import XGBClassifier

from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split, GridSearchCV
from sklearn.metrics import accuracy_score, classification_report, confusion_matrix
from sklearn.feature_selection import SelectKBest
from sklearn.feature_selection import chi2
from sklearn.decomposition import PCA
from sklearn.cluster import KMeans
from sklearn.datasets import make_classification
from sklearn.preprocessing import StandardScaler
from sklearn.impute import KNNImputer
from sklearn.feature_selection import mutual_info_classif

import warnings
warnings.filterwarnings("ignore")

df = pd.read_csv('cleaned_data.csv')
df.head(3)

# Function to create dataframe for columns and its unique count of values, datatypes and type of columns
def find_categorical_continous_variables(data, threashold):
    dic = {}
    continus_cols = []
    categorical_cols = []

    for column in data.columns:
        li = []
        li.append(len(data[column].unique()))
        li.append(data[column].dtype)

        # Check if the column is categorical or continuous
        if len(data[column].unique()) <= threashold or data[column].dtype == 'O':
            li.append("Categorical column")
            categorical_cols.append(column)
        else:
            li.append("Continous Column")
            continus_cols.append(column)

        dic[column] = li
    
    # create dataframe for columns and its details
    dic_df = pd.DataFrame(dic, index=["Unique values", "Data Type", "Categorical/Continous"])
    
    # return the dataframe, and lists for continous and categorical columns
    return dic_df.T, categorical_cols, continus_cols

# Function to find descriptive statistics by providing it the contious cols and dataframe
def descriptive_statistics_continous(df, continus_cols):
    dic={}
    for col in continus_cols:
        dic[col] = []
        dic[col].append(df[col].mean())
        dic[col].append(df[col].median())
        dic[col].append(df[col].mode()[0])
        dic[col].append(df[col].std())
        dic[col].append(df[col].var())
        dic[col].append(df[col].max() - df[col].min())
        dic[col].append(df[col].quantile(0.75) - df[col].quantile(0.25) )
        dic[col].append(df[col].skew())
        dic[col].append(df[col].kurtosis())

    df_details = pd.DataFrame(dic, index=["Mean", "Median", "Mode", "Std Deviation", "Variance", "Range", "IQR", "Skew", "Kurtosis"])
    return df_details.T

# lets find out the unique counts, datatypes, variable type like continous/categorical and lists for columns names having continous/categorical columns
col_type_df, categorical_cols, continus_cols = find_categorical_continous_variables(df, 10)

# display the data
col_type_df

# print columns 
print("Categorical Columns -\n",categorical_cols, "\n")
print("Continous Columns -\n",continus_cols)

Categorical Columns -
 ['status'] 

Continous Columns -
 ['name', 'MDVP:Fo(Hz)', 'MDVP:Fhi(Hz)', 'MDVP:Flo(Hz)', 'MDVP:Jitter(%)', 'MDVP:Jitter(Abs)', 'MDVP:RAP', 'MDVP:PPQ', 'Jitter:DDP', 'MDVP:Shimmer', 'MDVP:Shimmer(dB)', 'Shimmer:APQ3', 'Shimmer:APQ5', 'MDVP:APQ', 'Shimmer:DDA', 'NHR', 'HNR', 'RPDE', 'DFA', 'spread1', 'spread2', 'D2', 'PPE']

# Get the descriptive statistics 
descriptive_statistics_continous(df, continus_cols)

def plot_distribution_numerical(df):
    # Lets plot the histogram for each variable
    features = df.select_dtypes(include='number').columns

    # plot the subplot for histogram of each variable
    fig, axs = plt.subplots(23, 3, figsize=(20, 60))
    row = 0
    for feature in features:

        axs[row, 0].set_title("Histogram for {}".format(feature))
        sns.histplot(data=df, x=feature, kde=True, color="red", ax=axs[row, 0])

        axs[row, 1].set_title("Boxplot for {}".format(feature))
        sns.boxplot(data=df, x=feature,  color="skyblue", ax=axs[row, 1])

        axs[row, 2].set_title("Vaiolinplot for {}".format(feature))
        sns.violinplot(data=df, x=feature,  color="lightgreen", ax=axs[row, 2])
        row = row+1

    plt.title("Histogram, Boxplot and Violinplots for all variables")
    plt.tight_layout()
    plt.show()
    
plot_distribution_numerical(df)

---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
Cell In[8], line 24
     21     plt.tight_layout()
     22     plt.show()
---> 24 plot_distribution_numerical(df)

Cell In[8], line 10, in plot_distribution_numerical(df)
      7 row = 0
      8 for feature in features:
---> 10     axs[row, 0].set_title("Histogram for {}".format(feature))
     11     sns.histplot(data=df, x=feature, kde=True, color="red", ax=axs[row, 0])
     13     axs[row, 1].set_title("Boxplot for {}".format(feature))

IndexError: index 23 is out of bounds for axis 0 with size 23

# Covariance for pair of variables
def covariance_of_pair_variables(df):
    features = df.select_dtypes(include='number').columns
    dic_covariance = {}

    for index1 in range(0, len(features)):
        for index2 in range(index1+1, len(features)):
            dic_covariance[features[index1] + " - " + features[index2]] = df[features[index1]].cov(df[features[index2]])
            
    covariance_df = pd.DataFrame(dic_covariance, index=["Covariance"]).T
    return covariance_df.sort_values(by='Covariance')


covariance_of_pair_variables(df)

# Correlations matrix
plt.figure(figsize=(15, 15))
sns.heatmap(df.select_dtypes(include='number').corr(), annot=True, cmap='coolwarm')

<Axes: >

# plotting the scatterplots for all variables combinations 
sns.pairplot(df)

<seaborn.axisgrid.PairGrid at 0x1bb42c42dd0>

def t_test_for_two_groups(df, target_var, first_target, second_target):
    for feature in df.select_dtypes(include='number').columns:
        # Lets have two groups one for datapoints when "status is 0" and another when "status is 1"
        group_1 = df[df[target_var]==first_target][feature]
        group_0 = df[df[target_var]==second_target][feature]

        # Perform independent samples t-test
        t_statistic, p_value = ttest_ind(group_1, group_0)

        # Print the results
        print("\n---------{} for two outcomes of {}-------------\n".format(feature, target_var))
        print("T-statistic:", t_statistic)
        print("P-value:", p_value)

        # Interpret the results
        alpha = 0.05
        if p_value < alpha:
            print("Reject the null hypothesis: There is a significant difference in means of the two data groups of {} when {} is {} and {}".format(feature, target_var, first_target, second_target))
        else:
            print("Fail to reject the null hypothesis: There is NO significant difference in means of the two data groups of {} when {} is {} and {}".format(feature, target_var, first_target, second_target))
            
            
t_test_for_two_groups(df, 'status', 1, 0)

---------name for two outcomes of status-------------

T-statistic: -2.618996447591034
P-value: 0.009542284438269866
Reject the null hypothesis: There is a significant difference in means of the two data groups of name when status is 1 and 0

---------MDVP:Fo(Hz) for two outcomes of status-------------

T-statistic: -5.447839322812917
P-value: 1.592227189146496e-07
Reject the null hypothesis: There is a significant difference in means of the two data groups of MDVP:Fo(Hz) when status is 1 and 0

---------MDVP:Fhi(Hz) for two outcomes of status-------------

T-statistic: -2.105966006139477
P-value: 0.03654192080671688
Reject the null hypothesis: There is a significant difference in means of the two data groups of MDVP:Fhi(Hz) when status is 1 and 0

---------MDVP:Flo(Hz) for two outcomes of status-------------

T-statistic: -5.692141471654098
P-value: 4.784923852241762e-08
Reject the null hypothesis: There is a significant difference in means of the two data groups of MDVP:Flo(Hz) when status is 1 and 0

---------MDVP:Jitter(%) for two outcomes of status-------------

T-statistic: 4.0859747918809335
P-value: 6.506825054289377e-05
Reject the null hypothesis: There is a significant difference in means of the two data groups of MDVP:Jitter(%) when status is 1 and 0

---------MDVP:Jitter(Abs) for two outcomes of status-------------

T-statistic: 4.8761944285090975
P-value: 2.3037111641167024e-06
Reject the null hypothesis: There is a significant difference in means of the two data groups of MDVP:Jitter(Abs) when status is 1 and 0

---------MDVP:RAP for two outcomes of status-------------

T-statistic: 3.8768349749731166
P-value: 0.0001463942426461765
Reject the null hypothesis: There is a significant difference in means of the two data groups of MDVP:RAP when status is 1 and 0

---------MDVP:PPQ for two outcomes of status-------------

T-statistic: 0.6027388372559829
P-value: 0.5474126882565669
Fail to reject the null hypothesis: There is NO significant difference in means of the two data groups of MDVP:PPQ when status is 1 and 0

---------Jitter:DDP for two outcomes of status-------------

T-statistic: 3.856292612294128
P-value: 0.0001582581388863351
Reject the null hypothesis: There is a significant difference in means of the two data groups of Jitter:DDP when status is 1 and 0

---------MDVP:Shimmer for two outcomes of status-------------

T-statistic: 5.7197069054907725
P-value: 4.168890849126019e-08
Reject the null hypothesis: There is a significant difference in means of the two data groups of MDVP:Shimmer when status is 1 and 0

---------MDVP:Shimmer(dB) for two outcomes of status-------------

T-statistic: 5.5083786926335705
P-value: 1.1858552827842587e-07
Reject the null hypothesis: There is a significant difference in means of the two data groups of MDVP:Shimmer(dB) when status is 1 and 0

---------Shimmer:APQ3 for two outcomes of status-------------

T-statistic: 5.300447394450554
P-value: 3.2329358091596845e-07
Reject the null hypothesis: There is a significant difference in means of the two data groups of Shimmer:APQ3 when status is 1 and 0

---------Shimmer:APQ5 for two outcomes of status-------------

T-statistic: 5.5091649170951404
P-value: 1.1813090095950779e-07
Reject the null hypothesis: There is a significant difference in means of the two data groups of Shimmer:APQ5 when status is 1 and 0

---------MDVP:APQ for two outcomes of status-------------

T-statistic: 6.077737643304644
P-value: 6.6971916263062925e-09
Reject the null hypothesis: There is a significant difference in means of the two data groups of MDVP:APQ when status is 1 and 0

---------Shimmer:DDA for two outcomes of status-------------

T-statistic: 5.300281352950314
P-value: 3.2354923539354584e-07
Reject the null hypothesis: There is a significant difference in means of the two data groups of Shimmer:DDA when status is 1 and 0

---------NHR for two outcomes of status-------------

T-statistic: 2.647959426829298
P-value: 0.00878949536615006
Reject the null hypothesis: There is a significant difference in means of the two data groups of NHR when status is 1 and 0

---------HNR for two outcomes of status-------------

T-statistic: -5.409563773613764
P-value: 1.9161398150073093e-07
Reject the null hypothesis: There is a significant difference in means of the two data groups of HNR when status is 1 and 0

---------status for two outcomes of status-------------

T-statistic: inf
P-value: 0.0
Reject the null hypothesis: There is a significant difference in means of the two data groups of status when status is 1 and 0

---------RPDE for two outcomes of status-------------

T-statistic: 4.350125625164272
P-value: 2.2342434057464583e-05
Reject the null hypothesis: There is a significant difference in means of the two data groups of RPDE when status is 1 and 0

---------DFA for two outcomes of status-------------

T-statistic: 3.319606465498123
P-value: 0.0010836705446395163
Reject the null hypothesis: There is a significant difference in means of the two data groups of DFA when status is 1 and 0

---------spread1 for two outcomes of status-------------

T-statistic: 9.492051691133758
P-value: 1.0853373799969717e-17
Reject the null hypothesis: There is a significant difference in means of the two data groups of spread1 when status is 1 and 0

---------spread2 for two outcomes of status-------------

T-statistic: 7.067310887483291
P-value: 3.0381818141581874e-11
Reject the null hypothesis: There is a significant difference in means of the two data groups of spread2 when status is 1 and 0

---------D2 for two outcomes of status-------------

T-statistic: 5.3974150453986125
P-value: 2.0317665984241093e-07
Reject the null hypothesis: There is a significant difference in means of the two data groups of D2 when status is 1 and 0

---------PPE for two outcomes of status-------------

T-statistic: 8.695379609608944
P-value: 1.7455826911011754e-15
Reject the null hypothesis: There is a significant difference in means of the two data groups of PPE when status is 1 and 0

from sklearn import datasets, decomposition 
# Split the DataFrame into X and Y datasets

# Features (independent variables)
X = df.drop(columns=['status', 'name'])  

# Target variable (dependent variable)
y = df['status']

scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)

#Apply PCA
pca = PCA(n_components=2)
X_pca = pca.fit_transform(X_scaled)

#Combine PCA Components and Target Variable
df = pd.DataFrame(data=X_pca, columns=['PC1', 'PC2'])
df['Target'] = y

#Apply Clustering Algorithm
kmeans = KMeans(n_clusters=2, random_state=42)
df['Cluster'] = kmeans.fit_predict(df[['PC1', 'PC2', 'Target']])

# Visualize clustering results
import seaborn as sns
import matplotlib.pyplot as plt

plt.figure(figsize=(10, 6))
sns.scatterplot(data=df, x='PC1', y='PC2', hue='Cluster', style='Target', palette='viridis')
plt.title('PCA Clustering based on Target Variable')
plt.xlabel('Principal Component 1')
plt.ylabel('Principal Component 2')
plt.legend(title='Cluster')
plt.show()

	name	MDVP:Fo(Hz)	MDVP:Fhi(Hz)	MDVP:Flo(Hz)	MDVP:Jitter(%)	MDVP:Jitter(Abs)	MDVP:RAP	MDVP:PPQ	Jitter:DDP	MDVP:Shimmer	...	Shimmer:DDA	NHR	HNR	status	RPDE	DFA	spread1	spread2	D2	PPE
0	0	119.992	157.302	74.997	0.00784	0.00007	0.00370	0.00554	0.01109	0.04374	...	0.06545	0.02211	21.033	1.0	0.414783	0.815285	-4.813031	0.266482	2.301442	0.284654
1	1	122.400	148.650	113.819	0.00968	0.00008	0.00465	0.00696	0.01394	0.06134	...	0.09403	0.01929	19.085	1.0	0.458359	0.819521	-4.075192	0.335590	2.486855	0.368674
2	2	116.682	131.111	111.555	0.01050	0.00009	0.00544	0.00781	0.01633	0.05233	...	0.08270	0.01309	20.651	1.0	0.429895	0.825288	-4.443179	0.311173	2.342259	0.332634

	Mean	Median	Mode	Std Deviation	Variance	Range	IQR	Skew	Kurtosis
name	94.000000	94.000000	0.000000	54.703748	2.992500e+03	188.000000	94.000000	0.000000	-1.200000
MDVP:Fo(Hz)	154.605011	148.790000	88.333000	41.525291	1.724350e+03	171.772000	65.557000	0.592528	-0.633226
MDVP:Fhi(Hz)	198.026812	176.595000	102.145000	92.490876	8.554562e+03	489.885000	88.913000	2.515378	7.401087
MDVP:Flo(Hz)	115.305561	102.874000	1.000000	44.653702	1.993953e+03	238.170000	51.701000	1.130111	0.734658
MDVP:Jitter(%)	0.006093	0.004940	0.003690	0.004558	2.077129e-05	0.031480	0.003810	3.050555	12.379222
MDVP:Jitter(Abs)	0.000043	0.000030	0.000030	0.000034	1.157854e-09	0.000253	0.000030	2.775989	12.202619
MDVP:RAP	0.003236	0.002470	0.001690	0.002807	7.878800e-06	0.020760	0.002070	3.408734	15.450912
MDVP:PPQ	0.066817	0.002630	0.003320	0.872632	7.614869e-01	12.000000	0.002040	13.747551	188.996748
Jitter:DDP	0.009688	0.007420	0.005070	0.008418	7.085477e-05	0.062290	0.006180	3.420839	15.523735
MDVP:Shimmer	0.029264	0.022960	0.014500	0.017908	3.207078e-04	0.084650	0.021150	1.420081	1.657132
MDVP:Shimmer(dB)	0.282175	0.221000	0.154000	0.189523	3.591883e-02	0.933000	0.201000	1.704636	2.910687
Shimmer:APQ3	0.015492	0.012770	0.004690	0.009823	9.649054e-05	0.050830	0.012010	1.475233	2.208598
Shimmer:APQ5	0.017617	0.013430	0.007470	0.011290	1.274623e-04	0.049860	0.012740	1.446952	1.428675
MDVP:APQ	0.023583	0.017990	0.009030	0.015002	2.250460e-04	0.080890	0.016220	1.660942	2.946433
Shimmer:DDA	0.046476	0.038310	0.015670	0.029468	8.683737e-04	0.152510	0.036020	1.475234	2.208833
NHR	0.023823	0.011660	0.002310	0.038322	1.468603e-03	0.314170	0.018900	4.505491	25.951523
HNR	21.961164	22.085000	8.867000	4.268406	1.821929e+01	24.180000	5.763000	-0.445051	0.571981
RPDE	0.497594	0.491345	0.256570	0.103133	1.063635e-02	0.428581	0.163781	-0.129763	-0.909410
DFA	0.717117	0.722085	0.574282	0.054975	3.022233e-03	0.251006	0.086725	-0.039189	-0.663940
spread1	-5.725437	-5.736781	-7.964984	1.051205	1.105032e+00	5.125228	1.320026	0.347205	-0.210744
spread2	0.223666	0.217013	0.210279	0.081229	6.598211e-03	0.428052	0.099760	0.074474	-0.130574
D2	2.376950	2.361532	1.423287	0.368745	1.359730e-01	1.990362	0.531619	0.312325	-0.072176
PPE	0.203052	0.192730	0.044539	0.086461	7.475510e-03	0.412994	0.113350	0.721347	0.261612

	Covariance
name - MDVP:Flo(Hz)	-143.088761
name - HNR	-48.473436
MDVP:Fo(Hz) - spread1	-19.036930
MDVP:Flo(Hz) - spread1	-18.886935
MDVP:Flo(Hz) - status	-7.539918
...	...
name - MDVP:Fo(Hz)	81.688888
MDVP:Fhi(Hz) - MDVP:Flo(Hz)	351.507050
name - MDVP:Fhi(Hz)	706.529244
MDVP:Fo(Hz) - MDVP:Flo(Hz)	1081.942775
MDVP:Fo(Hz) - MDVP:Fhi(Hz)	1511.889706

Importing Libraries¶

Importing Cleaned Data¶

Data Exploration (EDA)¶

Univariate analysis on numerical data¶

Descriptive Statistics¶

Conclusion -¶

Conclusion -¶

Conclusion -¶

Bivariate analysis on numerical data¶

Conclusion -¶

Conclusions -¶

Multivariate analysis on numerical data¶

	Unique values	Data Type	Categorical/Continous
name	189	int64	Continous Column
MDVP:Fo(Hz)	189	float64	Continous Column
MDVP:Fhi(Hz)	189	float64	Continous Column
MDVP:Flo(Hz)	189	float64	Continous Column
MDVP:Jitter(%)	168	float64	Continous Column
MDVP:Jitter(Abs)	18	float64	Continous Column
MDVP:RAP	149	float64	Continous Column
MDVP:PPQ	162	float64	Continous Column
Jitter:DDP	174	float64	Continous Column
MDVP:Shimmer	182	float64	Continous Column
MDVP:Shimmer(dB)	145	float64	Continous Column
Shimmer:APQ3	180	float64	Continous Column
Shimmer:APQ5	183	float64	Continous Column
MDVP:APQ	183	float64	Continous Column
Shimmer:DDA	184	float64	Continous Column
NHR	181	float64	Continous Column
HNR	189	float64	Continous Column
status	2	float64	Categorical column
RPDE	189	float64	Continous Column
DFA	189	float64	Continous Column
spread1	189	float64	Continous Column
spread2	188	float64	Continous Column
D2	189	float64	Continous Column
PPE	189	float64	Continous Column